Link Search Menu Expand Document

Project Publications

Chapter 4: Conclusions: Legal Protection and its Dependencies on Text-driven Normativity

By Mireille Hildebrandt

This Research Study is part of the output of the COHUBICOL research project that is focused on fundamental research into computational ‘law’. The project asks whether and under what conditions ‘legal techs’ could and/or should qualify as ‘law’ (i.e. understood as legitimately generating legal effect) and the subsequent question of how it may directly or indirectly impact law (i.e. directly or indirectly affect legal effect). The answer to these questions should serve the aim of the project, that is to investigate how we can make sure that human beings who are subject to a law that is built on ‘legal techs’, actually ‘count’ as human agents.

The meaning of ‘counting as’ integrates both the quantitative meaning of counting (one person one vote) and its qualitative meaning (qualifying as)

For humans to ‘count as’ a human being we need to foster (1) the checks and balances of the Rule of Law that requires a government to treat each and every citizen with equal respect and concern and (2) a democratic rule based on one person one vote, thus allowing for a majority rule. However, in the context of constitutional democracies, i.e. democracies that respect the Rule of Law, the ruling majority must ensure equal respect and concern for those who are not part of that majority. This is where populism as well as authoritarian rule miss the mark, overruling individual agency once a majority has been formed or a ruler installed. The meaning of ‘counting as’ thus integrates both the quantitative meaning of counting (one person one vote) and its qualitative meaning (qualifying as). Note that in this study we do not use the concept of ‘liberal democracy’ as this has too many connotations that link democracy to capitalism as intrinsically conducive to liberalism. Instead, we use the term constitutional democracy to emphasise that a viable democracy depends on the checks and balances of the Rule of Law, where the latter concept is less about ‘rules’ (legalism) than about law as the binding normative framework that restricts the decision space of the government (legality). The concepts of ‘liberal democracy’ and ‘Rule of Law’ derive from the Anglo-American discourse, whereas ‘constitutional democracy’ and Rechtsstaat or Etat de Droit are more akin to continental framings; it would be great if the discourse of political theory, Rule of Law and democracy would take into account more of continental philosophy instead of requiring whoever writes in English to first explain how their analyses can be expressed in the vernacular of analytical philosophy and either utilitarianism or deontological moral philosophy.

The project thus takes a specific normative position on the Rule of Law, highlighting its core as the antinomian interplay between legal certainty, justice and instrumentality (Radbruch), while asserting that legal certainty entails both the contestability of legal norms and their final authoritative interpretation by an independent court (Waldron). Roger Brownsword has convincingly argued that a law that cannot be disobeyed is not law but discipline, management or administration. We add that norms that cannot be contested in a court of law are not legal norms but discipline, management or administration, and should therefore be open to contestation as such under the Rule of Law. This normative position implies that law is an argumentative practice rather than the application of either logic or statistics, thus raising a number of flags around attempts to ground the law on technologies that necessarily reduce legal decision-making to the application of logic and/or statistics.

Law is an argumentative practice rather than the application of either logic or statistics

The goal of this ‘Research Study on Text-Driven Law’ is to probe the added value of the three framing concepts that are key to the COHUBICOL project with regard to modern positive law, understood as text-driven law. These three concepts are (1) the ‘mode of existence’ of law, (2) ‘affordances’, i.e. law as an ‘affordance’ and the ‘affordances’ of law and (3) ‘legal protection by design’.

4.2 Chapter 1: the text-driven nature of modern positive law

Before engaging with the triple framing concepts, chapter 1 prepares the ground by explaining how the information and communication infrastructure (ICI) of the printing press afforded the key principles of the Rule of Law, notably the contestability of legal decision-making combined with the need for closure by an independent court. Chapter 1 recounts how the rise of the ICI of the printing press turned the need for interpretation into the hallmark of modern positive law. This was due to the distance in time and space that is created by printed text: distance between an author (e.g. a legislature) and its audience (e.g. those who share jurisdiction), between an author and their text (legislation), due to the spatial reach of printed text compared to both handwritten and oral speech, thus also creating distance between the author and the meaning of their text (in concrete legal decisions), due to the loss of control over how the reader will interpret the text, e.g. because of the death of the author (after elections a new author, i.e. a new legislature, takes office). It was this need for interpretation that triggered both the need for closure and the inevitable ability to contest, inviting a balancing act that steers free from both rigid rule application (legalism) and arbitrary rule (Einzelfallgerechtigkeit). The point is that this delicate balancing act cannot be taken for granted; it depends on the institutionalisation of countervailing powers. The proliferation of the ICI of the printing press did not ‘cause’ or ‘logically imply’ the rise of the Rule of Law, it merely afforded the Rule of Law as an institutionalised practice.

4.3 Chapter 2: the framing concepts

This chapter grounds the framing concepts of the project (mode of existence, affordance and legal protection by design) in the works of the project’s author. It brings together a series of seminal quotes from her book Smart Technologies and the End(s) of Law and from the project proposal ‘Counting as a human being in the era of computational law’ that directs the research. Below, I reiterate the working definitions of the framing concepts, as related in chapter 2, clarifying both differences and similarities with other ways of using them.

4.3.1 The modes of existence (MoE) of modern positive law

The concept of a MoE was introduced in the Project as a way to highlight that modern positive law exists in a specific way, compared to other types of legal traditions (e.g. medieval, Roman, religious) and compared to other societal domains (notably morality and politics, but also economics or religion). Core to the idea of MoE is that speaking and writing can be ways of acting, bringing about performative effects while creating so-called institutional facts. Modern positive law is the prime example of such acts, exemplified in the notion of legal effect, that is neither caused nor logically inferred but constituted by speech acts such as: enacting legislation, concluding a contract, deciding a judgment.

Current law’s mode of existence is an affordance of the technology of text, which in turn also afforded the institutional checks and balances of the Rule of Law

The concept was inspired by Latour’s usage, which was in turn inspired by Souriau and similarly inspired by Stengers’ concept of an ecology of practices. In Smart Technologies and the End(s) of Law, however, Hildebrandt developed her own conception of the way that law-as-we-know it exists, highlighting the relationship between, on the one hand, modern positive law and the Rule of Law and, on the other hand, the information and communication infrastructure (ICI) of the printing press. A key difference may be that she argues that current law’s mode of existence is an affordance of the technology of text, which in turn also afforded the institutional checks and balances of the Rule of Law. The Project is based on the assumption that we cannot presume that once law becomes grounded in another ICI its affordances in terms of legal protection will remain the same, more notably with regard to the legal protection offered under the Rule of Law. The Project aims to investigate how this will affect law’s current mode of existence, more notably the nature of legal effect and related institutional foundations.

The use of this concept should thus enable us to become aware of the fact that modern positive law’s mode of existence cannot be taken for granted once the text-driven ICI is integrated with code- and/or data-driven ICIs, while allowing us to acknowledge that this will transform the mode of existence of law-as-we-know-it, requiring deliberate efforts to secure the kind of contestation and protection that is key to the Rule of Law.

4.3.2 Affordance of and affordance for

The concept of affordance, as used in this Project, builds on Gibson’s original concept that was part of his ecological psychology, which traced the relational nature of what are often considered ‘properties’ of a specific environment:

The affordances of the environment are what it offers the animal, what it provides or furnishes, either for good or ill. The verb to afford is found in the dictionary, but the noun affordance is not. I have made it up. I mean by it something that refers to both the environment and the animal in a way that no existing term does. It implies the complementarity of the animal and the environment.

The concept has been further developed by Norman in the domain of design, initially focused on how to design technological artefacts in such a way that the intended affordances of a product are more easily perceived as such by the relevant user. From the perspective of the Project we are however also interested in what hidden affordances can be manipulated by the provider or deployer of a system to influence end-users, e.g. without their conscious awareness.

Whereas Gibson was mostly focused on the affordances of the material environment, we focus on both the material environment (for instance specific technologies) and the institutional environment (created and sustained by performative speech acts that may be mediated by specific technological infrastructures). We understand the affordances of the material & institutional environment of human beings as what the material & institutional environment offers us as embodied human agents. Note that this does not imply a mutually exclusive conceptualisation of ‘material’ and ‘institutional’; institutional facts are afforded by the ‘matter’ of our own embodiment (our voice, brains, gestures, mobility) and the affordances of our material environment are determined by institutional facts (created on the cusp of language, human agency and real-world navigation). With the term affordance we thus refer to the relation between a material & institutional environment on the one hand and the human agents that find themselves in that environment while also shaping it on the other hand. Instead of defining law as a property of human society the concept of an affordance foregrounds the inbetween and mutual dependencies of human agents and their environments.

Law as an ‘affordance of’: this understanding of affordance allows us to pay keen attention to the affordances of a specific information and communication infrastructure (ICI) for the constitution of the law, acknowledging that law-as-we-know-it is an affordance of a text-driven ICI.

Law as an ‘affordance for’: this, in turn, allows us to understand legal protection as an affordance of modern positive law that forms the material & institutional environment of natural persons (human beings), acknowledging that law-as-we-know-it has specific affordances for human beings.

Modern positive law then, should be understood as a system of legal written speech acts and their resulting institutional facts, built on the ICI of text. This way of seeing the relationship between the material and the institutional environment of law and the relationship between specific material & institutional environments and human agents, will allow us to compare the affordances of a text-driven ICI to those of code- or data-driven ICIs, particularly as to legal protection.

4.3.3 Legal protection by design (LPbD) in the era of text-driven normativity

LPbD is a term coined by Hildebrandt to refer to the articulation of legal protection into the prevailing information and communication infrastructure (ICIs), more notably the legal protection provided by fundamental rights and the institutional checks and balances (countervailing powers) of the Rule of Law. LPbD is not equivalent with Lawrence Lessig’s ‘Code as Law’, which frames the normative force of computing code in terms of ‘architecture’, next to social, economic and legal norms. This Project is based on the understanding that social, economic and legal norms overlap in various ways, thus also highlighting that the extent to which computer code determines human behaviour depends on the affordances of the relevant computing systems.

LPbD should not be confused with techno-regulation, which refers to both legal and non-legal, and — in case of the latter — both deliberate and unintended regulatory effects of technologies. Based on the understanding that ‘technology is neither good nor bad, but never neutral’, technologies have normative affordances that may be part of deliberate design decisions, aimed to have specific intended effects, though such normative affordances may also be unintended ‘side-effects’.

Legal Protection by Design must be distinguished from ‘ethics by design’ or ‘values by design’

LPbD must also be distinguished from ‘ethics by design’ or ‘values by design’, which is based on the acknowledgement that any design will have normative and possibly moral implications, inevitably embedding certain values, whether or not the designer is aware of this. LPbD aims to incorporate the specific values of fundamental rights and the checks and balances of the Rule of Law into prevailing ICIs, grounding the design in democratic participation and legislation while ensuring contestability as a core and actionable value of legal protection. LPbD is not about embedding a designer’s ethical preferences into a product or infrastructure but about protection against intended and unintended disruption of legal effect due to the use of legal technologies.

LPbD should also not be confused with ‘legal by design’, which refers to a specific type of techno-regulation, whereby legal norms are translated into code or into the design of computing systems such that compliance become automated or semi-automated. Think of self-executing code as in smart contracts or smart regulations, or data-driven techniques for prediction of judgments deployed to make legal decisions. Legal by design aims to attribute legal effect to the operations of technologies, thus automating the attribution of legal effect which runs counter to the protection of the Rule of Law.

The concept of LPbD will allow us to investigate to what extent, and if so how, legal protection as an affordance of a text-driven ICI can be re-articulated in code- and/or data-driven ICIs.

Together, these three framing concepts highlight the relational and ecological nature of the law, literally foregrounding the ‘texture’ of modern positive law.

4.4 Chapter 3: the conceptual scaffolding of modern positive law

This chapter is based on the initial phase of the project, which consisted of an in-depth investigation of a set of foundational legal concepts in the light of the framing concepts. The chapter introduces a set of fundamental concepts that are constitutive for law and the Rule of Law, while situating them in relation to each other. After this, the chapter explains each concept in terms of the framing concepts, thus exploring the added value of ‘mode of existence’, ‘affordance’ and ‘legal protection by design’ as a way to provide a better understanding of the law.

The materiality of text affords abstract thought and an external memory that calls for reiterant interpretation, thus affording both contestation and closure

Based on the discussion of foundational legal concepts in chapter 3, the concept of a ‘legal norm’ can be situated in relation to the other foundational legal concepts as follows:

  • Legal norms determine a specific legal effect when certain conditions are fulfilled. Their validity, and thus their legal effect, depends on the relevant jurisdiction (not on individual preferences, moral agreement or economic power). Legal effect can, for instance, be the transfer of property, the obligation to pay a price, the legal power to punish a person or the legal power to impose the payment of compensation; a legal effect can also be ‘that a certain action is lawful’. Subjective rights are created by legal norms that attribute such rights. The delivery of a good based on a valid contract of sale creates a subjective right with regard to that good, more precisely a property right. This is based, in turn, on legislation or case law that attributes the legal effect of ‘having a property right’ to the performance of a contract. Legal effect is a performative effect in that it does what it says. To have that specific performative effect, legal norms must be derived from the sources of law: international treaties, the constitution, legislation, case law, doctrine, custom and fundamental principles of law. Legal norms must be distinguished from physical laws and from moral obligations or threats. Legal norms can in principle be violated (physical laws cannot) and the obligation to comply with them does not depend on the individual moral agreement of those subject to law or on the arbitrary wishes of a government official. Some legal norms prescribe or prohibit specific actions (primary or regulative norms), other legal norms decide who have the legal power to enact, change or adjudicate the interpretation of legal norms (secondary or constitutive norms). Deciding the meaning of a legal norm in a concrete case requires legal reasoning and legal interpretation; rules and principles do not interpret themselves. Though legal norms take the form of a rule their interpretation depends on binding fundamental legal principles such as ‘good faith’ or ‘proportionality’. Legal norms address legal subjects, that is both natural persons (human beings) and entities that have been given legal personhood (e.g. corporations). Legal norms may attribute subjective rights to legal subjects, based on the relevant positive law. In a constitutional democracy, legal norms both constitute and limit the legal powers of the state, thus bringing the government under the Rule of Law.

Based on the discussion of the framing concepts in chapter 2, the concept of a ‘legal norm’ can be understood as follows:

  • The mode of existence of legal norms in the era of modern positive law is constituted by the legal effect they generate and this legal effect is a performative effect that is afforded by spoken and written speech acts. The materiality of that mode of existence can be found in (1) the embodiment of natural language and (2) the embedding of written law in the technologies of text. The materiality of the human body affords the development of complex human languages and speech acts that create an institutional environment. The materiality of text affords abstract thought and an external memory that calls for reiterant interpretation, thus affording both contestation and closure. The performative effect of spoken and written speech acts defines the legal protection that is offered by modern positive law. One could say that in this sense, the kind of legal protection offered by law-as-we-know-it is contingent upon its articulation in spoken and written language. The design of these speech acts determines their performative effect and in that sense the enactment of legal norms by the legislature and the courts can be seen as a matter of legal protection by design.

4.4.2 Rule of law and positive law

Based on the discussion of foundational legal concepts in chapter 3, the concept of the ‘Rule of Law’ can be situated in relation to the other foundational legal concepts as follows:

  • The Rule of Law (état de droit, Rechtsstaat) refers to the institutionalisation of checks and balances within the state, making sure that countervailing legal powers keep each other in check, thus preventing the arbitrary exercise of public power over legal subjects, notably over ‘natural persons’.

The Rule of Law thus limits the legal power to attribute of legal effect in a way that offers practical and effective legal protection to legal subjects, especially natural persons, by attributing them subjective rights. The difference between the Rule of Law and a rule by law refers to the difference between, on the one hand, a law that is both an instrument of public policy (creating legal effect) and an instrument of protection (simultaneously limiting the attribution of legal effect) and, on the other hand, a law that is nothing but an instrument to achieve public policy goals as defined by the government. The Rule of Law implies legality, meaning that state powers can only be exercised within the bandwidth of legal powers attributed for specified and legitimate purposes, taking into account human rights while respecting independent judicial review. Rule by law may refer to legalism, where state powers can be and must be exercised in accordance with the will of the legislator and/or to legal formalism, where the state has discretionary powers to achieve their objectives as long as these powers have been attributed in accordance with specified procedures.

Even though the Rule of Law may apply to different jurisdictions, each jurisdiction will articulate the checks and balances in different ways. The Rule of Law depends on the interpretation of the sources of law in a specific jurisdiction, more precisely, the Rule of Law stipulates that such interpretation must be done by an independent judiciary. The sources of law are defined in the context of specific jurisdictions, both national and international, thus linking the Rule of Law to democratic law-making. The sources of law define positive law within a specific jurisdiction, thus embedding the Rule of Law in concrete legal orders. For positive law to qualify as articulating the Rule of Law, however, the interpretation of legal norms derived from the sources of law must be in the hands of an independent judiciary. Legal reasoning, then, involves the anticipation of how a court will interpret a legal norm in the context of positive law in the relevant jurisdiction.

In Anglo-American legal philosophy Rule of Law is often equated with conditions such as accessibility, clarity, generality, non-contradiction, non-retroactive application, feasibility and foreseeability, coupled with the notion of an independent judiciary (Fuller). A difference is often made between a thin and a thick version, depending on whether conditions are more formal or more substantive. In the latter case more attention is given to human rights protection, including social and cultural rights. Others, however, pay keen attention to rights of contestation against the state (Dicey), and to procedural conditions that enable contestation and argumentation as core to the Rule of Law (Waldron), and to formal characteristics that can constrain what a legitimate legal rule can possibly be (Wintgens).

In continental European legal theory, the Rechtsstaat or Etat de Droit can similarly be seen in a more formal or substantive way, with keen attention to the extent to which the legal powers of the state are limited, including the question of whether states have positive obligations to ensure respect for human rights in both the public and the private sphere. The Rule of Law thus entails the protection of subjective rights that cannot be overruled without due process of law.

In the context of COHUBICOL we take a substantive and procedural perspective on the Rule of Law, integrating a formal perspective in a way that embraces legality while rejecting both legalism and arbitrary rule, incorporating ‘practical and effective’ protection of human rights and access to an independent court to ensure the contestability of actions or decisions in the public or private sphere that may violate rights or obligations.

Based on the discussion of foundational legal concepts in chapter 3, the concept of ‘positive law’ can be situated in relation to the other foundational legal concepts as follows:

  • Positive law is the entirety of legal norms, derived from the sources of law that are in force in a specific jurisdiction, at a specific point in time. As explained under legal norms, this includes both primary rules (regulative, i.e. legal norms that directly regulate) and secondary rules (constitutive, i.e. legal norms that define how primary rules can be made) Being in force refers to (1) the binding character of positive law, (2) the state’s actual power to enforce the law and (3) a decision by a legislator, public administration or court whereby they enact legal norms in the sense of issuing, interpreting and/or applying them. All three points relate to the nature of legal effect as opposed to causal effect or logical inference. Legal certainty depends on the ‘positivity’ of the law. Positive law is informed by the moral principles that constitute its implied philosophy and simultaneously informs the moral practices of those subject to its normativity. Positive law differs from morality in (1) that it does not depend on the moral inclinations of an individual decision-maker, and (2) that it is in principle enforceable against those under its jurisdiction. Positive law differs from politics and policy in that it does not determine the purposes of a polity but determines what legal effect is attributed based on the fulfilment of what legal conditions. The Rule of Law implies that political decision-making depends on the attribution of a legal power to do so, meaning that the legal effect of primary legal norms depends on the legal effect of secondary legal norms.

Positive law assumes the existence of a sovereign state and simultaneously constitutes and regulates that same sovereign state. The Rule of Law as well as the protection of human rights depend on positive law. Positive law is often opposed to ‘natural law’, which may refer to divine law (medieval period) or the law of reason (enlightenment period), both of which claim universal application and an objective truth-value; positive law is human-made (it is ‘posited’), depending on the social contract that defines a particular jurisdiction. Though some authors restrict the meaning of ‘positive law’ to legislation, we use the concept to refer to all legal norms, whether enacted by a legislature or a court, whether written or unwritten, as long as they derive from the sources of law.

By thinking in terms of affordances rather than causation or logical inference, we can avoid technological determinism, while nevertheless demonstrating what is made possible by a law that depends on the ICI of the proliferating printed text

Positive law should not be confused with ‘legal positivism’, which refers to a specific conception about the nature of law, its making and its validity. Recognizing the importance of positive law does not imply ‘legal positivism’.

Based on the discussion of the framing concepts in chapter 2, the concepts of ‘Rule of Law’ and ‘positive law’ can be understood as follows:

  • The mode of existence of the Rule of Law and positive law implies that they are co-constitutive. In the words of Gori ‘on the one hand, the “law which rules” is positive law; on the other, the constellation of values and normative standards enshrined by the doctrine of the Rule of Law informs the understanding of what is required for positive law to “rule”.’ The normative force of both positive law and the Rule of Law hinge on the text-driven normativity that informs written and unwritten law (taking note that unwritten law only makes sense in the context of written law). The affordances of natural language and the infrastructure of the printing press allowed for the rise of both positive law and the Rule of Law, because of the inherent multi-interpretability of language and the ensuing contestability of printed text. By thinking in terms of affordances rather than causation or logical inference, we can avoid technological determinism, while nevertheless demonstrating what is made possible by a law that depends on the ICI of the proliferating printed text. This also asserts that the kind of legal protection that is afforded by positive law under the Rule of Law, is in turn an affordance of a text-driven ICI. One could say that this text-driven ICI, informing both positive law and the Rule of Law, offers a kind of legal protection by design, even if this does not necessarily entail a deliberate attempt on the side of those who invented the printing press.

Based on the discussion of foundational legal concepts in chapter 3, the concept of ‘legal effect’ can be situated in relation to the other foundational legal concepts as follows:

  • Legal effect is the consequence of a legally relevant fact, where that consequence is attributed by positive law and consists of a change in the legal status of a legal subject, including a change in their legal powers, their subjective rights or obligations.

The attribution of legal effect can entail e.g. the attribution of a right or legal power, the voiding of an obligation, or the qualification of some state or behaviour as either lawful or unlawful. The attribution of legal effect is brought about by a legal norm that consists of a set of legal conditions (Tatbestand) that attribute the legal effect if the conditions are fulfilled. The attribution is neither caused nor logically inferred; it is performative in the sense of speech act theory. For instance, fulfilling the conditions that constitute a criminal offence has the legal effect of being punishable, not of being punished. This clarifies that the effect is performative and not causal; being punishable is an institutional fact whereas being punished would be a brute fact.

The set of legal conditions (Tatbestand) that result in a legal effect are specified in positive law, more precisely in a source of law: treaties, legislation, case law, customary law, or fundamental principles of law. As positive law depends on the relevant jurisdiction, legal effect in turn differs per jurisdiction, even if some legal effects may apply in many jurisdictions. The decision on legal effect is first made by the legislature that may enact new sources of law (legislation), but under the Rule of Law the last word is with an independent court, based on explicit legal reasoning that supports the court’s interpretation of the relevant legal norm.

Based on the discussion of foundational legal concepts in chapter 3, the concept of ‘sources of law’ can be situated in relation to the other foundational legal concepts as follows:

  • The sources of law refer to the set of written and unwritten resources from which binding legal norms are ‘drawn’; the sources do not contain information about the law, they constitute the law as they define what counts as law.

The sources of law are usually limitatively summed up as: international treaties, legislation, case law and doctrine (written sources) and fundamental principles of law and customary law (unwritten sources). All these written and unwritten sources present the binding legal norms that define what legal effect is to be attributed depending on what conditions, within a specific jurisdiction. A constitution can be written (legislation) or unwritten (customary law); either way, it constitutes the legal powers of the state (as a legal person) and the fundamental rights of its citizens (as legal subjects). The content of the sources of law will differ per jurisdiction and will depend on positive law.

Doctrine contributes to the interpretation of binding legal norms, though it is not binding in itself, the same goes for recitals in treaties, opinions of advocats general (advisors) of highest courts and other formal advisory bodies (e.g. the European Data Protection Board). The binding force of fundamental legal principles does not depend on whether or not and how they have been codified in written sources; they are tied up with the core tenets of the Rule of Law and the moral and institutional grounding of the law. Customary law binds due to usus (actual adherence) and opinio necessitatis (a shared sense of obligation).

To select and apply a relevant legal norm implies an act of interpretation; the act of selection and application cannot be reduced to a logical sequence though it must be justifiable in the form of a syllogism; the need to justify the choice and the interpretation of a legal norm restricts the decisional space of public administration and the courts, thus bringing them under the Rule of Law. Interpretation cannot be arbitrary, legal doctrine distinguishes grammatical, systematic, historical and teleological interpretation, i.e. taking into account the ordinary meaning of the relevant terms, the place of the norm within the relevant legal source, the legislature’s intent as derived from official documents, and the aims of the relevant legal source. This implies that interpretation requires legal reasoning, which it also affords. Though courts have discretion in selecting and combining these methods of interpretation, the exercise of such discretion is bounded by the demands of legal certainty, justice and the purposiveness of the law.

Based on the discussion of foundational legal concepts in chapter 3, the concept of ‘jurisdiction’ can be situated in relation to the other foundational legal concepts as follows:

  • Jurisdiction refers to the legal powers of a legal order and to the scope of such power. This means that jurisdiction defines what legal norms constitute positive law based on the sources of law of the relevant jurisdiction. Jurisdiction thus also determines the attribution of subjective rights and the distribution of legal powers.

Jurisdiction may refer to: a sovereign’s legal powers to legislate, adjudicate and enforce; the territory or domain over which a sovereign those legal powers; the competence of a specific court to adjudicate, which is defined by a combination of material and procedural conditions. Since the Peace of Westphalia (1648) jurisdiction depends on sovereignty, which in turn is defined by territorial jurisdiction. This circular interdependence relates to two sides of the same coin: internal sovereignty provides for national jurisdiction and vice-versa while external sovereignty defines international jurisdiction and vice-versa. Internal sovereignty cannot exist without external sovereignty and vice-versa.

Jurisdiction can in principle be based on territory (modern positive law is aligned with territorial jurisdiction); personal status (birth, kinship, membership of a religion); subject matter (criminal jurisdiction, private law jurisdiction) or on the effect of an action that gives rise to a legal claim (e.g. in tort law). In all cases, jurisdiction is about what institution decides the attribution of legal effect within its domain (whether based on territory, personal status or subject matter) as well as which legal subjects are subject to its binding force. It thus also decides what subjective rights are attributed under what conditions.

The mode of existence of legal effect, the sources of law and jurisdiction is a ‘matter’ of institutional facts, that is – they are the performative effect of a dynamic network of written and spoken speech acts

In the current world order, we can distinguish national, international and supranational jurisdictions. As to national jurisdiction we can distinguish: internal jurisdiction, that is, the competence to legislate, adjudicate, and enforce the law within the state; extraterritorial jurisdiction, that is, the competence of one state to legislate, adjudicate, or enforce its law in relation to legal subjects, actions or effects in the territory of another state. International jurisdiction depends on the sources of international law. The relationship between potentially overlapping jurisdictions is itself subject to the jurisdiction of a national court (e.g. international private law) or an international court (notably in international public law). The question who gets to decide on jurisdiction is often called: Kompetenz-Kompetenz; it refers to the question of what entity has jurisdiction to decide jurisdiction. This question is key for decisions on the applicability and interpretation of legal norms, especially where different bodies claim jurisdiction concerning the same territory, persons or subject matter. Legal reasoning depends on the ability to decide what sources of law apply, what legal norms are part of positive law, thus enabling the interpretation of a legal norm in the context of the whole of the applicable legal norms.

Based on the discussion of the framing concepts in chapter 2, the concepts of ‘legal effect’, ‘sources of law’ and ‘jurisdiction’ can be understood as follows:

  • The mode of existence of legal effect, the sources of law and jurisdiction is a ‘matter’ of institutional facts, that is – they are the performative effect of a dynamic network of written and spoken speech acts. The networked character of both natural language and its usage, on the cusp of intra- and extra linguistic reference, accords with the fact that legal norms and legal effect must always be situated in the whole of the sources of law that are applicable within a specific jurisdiction. In that sense language and language use – seen as acts of embodied human agents in a physical and institutional environment – afford jurisdiction and the sources of law as ‘enlanguaged affordances’. As Van den Hoven writes in chapter 3: ‘This notion [enlanguaged affordances] puts emphasis on the ways in which the affordances of the human ecological niche are interwoven with practices of speaking and writing. Speech and writing allow us to engage with affordances across long timescales and allow us to think in abstract and institutional terms about the world. (…) The very nature of the web of meaning, as Taylor puts it, is to be “present as a whole in any one of its parts. To speak is to touch a bit of the web, and this is to make the whole resonate”.’ This is also what makes legal protection possible, ensuring that legal norms are not decided in splendid isolation but in the context of the sources of law that inform the relevant jurisdiction. As Van den Hoven clarifies, this entails that legal protection by design is not a matter of rejecting technologies other than those of the word, ‘rather it can be understood as a manifesto for the preservation of thoughtfulness in law’. Even if the legal protection as-we-know-it was not the intended result of a text-driven ICI, that ICI nevertheless has a certain design that is conducive to both the contestability and the closure that are key to jurisdiction, the sources of law and legal effect. In that sense legal protection by design is core to text-driven law, notably to its networked and systemic dimensions, as visible in the notions of ‘sources of law’ and ‘jurisdiction’.

Based on the discussion of foundational legal concepts in chapter 3, the concept of ‘legal subject’ can be situated in relation to the other foundational legal concepts as follows:

  • A legal subject is an entity capable of acting in law, of having subjective rights and legal obligations. Most jurisdictions attribute legal subjectivity to two types of entities: to all human beings (called natural persons) and to any other entity qualified as such (called legal persons) by the legislature or the courts. This clarifies that it is positive law that decides whether and under what conditions an entity is a legal subject. Corporations, ships, trees or AI systems may qualify as legal persons, depending on what legislatures or courts within a specific jurisdiction decide. As positive law depends on the sources of law that apply in a specific jurisdiction, the attribution of legal subjectivity may differ amongst jurisdictions.

In current constitutional democracies, natural persons have full legal subjectivity (in all domains of law), whereas legal persons have restricted legal subjectivity, as defined by the relevant positive law. In most jurisdictions the following entities are given legal personhood: the state and public bodies such as cities or regions, international organisations, corporations (various types) and associations, foundations or charities. Legal persons will necessarily require representation by one or more natural persons to act in law, to exercise their standing in court, to exercise their rights and to fulfil their legal obligations. Legal personhood is restricted to the remit defined by the legislator or the courts, which means they are not necessarily entitled to human rights. The extent to which legal persons can be held liable in private, public or criminal law and/or exercise specific rights is not only a matter of legislative attribution but may also be a matter of interpretation. The European Court of Human Rights (ECtHR) has e.g. decided that corporations may have a restricted right to privacy. This allows the national courts of the Council of Europe to develop legal reasoning that fits with the interpretation of the ECtHR.

Being a legal subject enables an entity to generate legal effect, whether intended (as in the case of a contract) or unintended (as in the case of a tort). The ability to generate intended legal effects provides legal subjects with legal powers, e.g. the power to conclude a valid contract. The attribution of legal subjectivity implies the attribution of subjective rights; in the case of natural persons, those rights include human or fundamental rights, e.g. privacy, non-discrimination, fair trial or the right to an effective remedy to invoke one’s rights.

Legal subjects may have limited capacity, as defined by positive law, e.g. minors may not enter contracts, unless authorised by their parents, minors may not be liable under tort law, though their parents may be liable instead, corporations may be able to conclude contracts and be held liable under private law, but may not be punishable under criminal law, and natural persons may be placed under guardianship in case of mental incapacity, in which case they cannot perform juridical acts (i.e. acts with intended legal effect). Whether and under what conditions legal subjects have limited capacity depends on positive law and thus on the relevant jurisdiction.

A legal subject is not the same thing as either the human being or the non-human entity that is granted legal subjectivity. Instead, it is akin to an avatar that enables them to play specific role(s) in law. This means that being a legal person does not imply being a moral person, that is a being capable of acting morally.

Based on the discussion of foundational legal concepts in chapter 3, the concept of ‘subjective rights’ can be situated in relation to the other foundational legal concepts as follows:

  • A subjective right is always relational (between legal subjects, with regard to one or more legal objects, such as a property or an obligation). It can be one or more of the following: a claim – attributed by positive law – of a legal subject, that one or more other legal subjects act or do not act in a certain way in relation to that legal subject, and/or a liberty – attributed by positive law – of a legal subject, that they are free to act in a certain way in relation to one or more other legal subjects, and/or a legal power – attributed by positive law – of a legal subject, that they are free to use in relation to one or more other legal subjects. A subjective right is the legal effect of a specific legal norm, derived from the relevant sources of law, within a specific jurisdiction. Legal subjects have the legal power to create or transfer subjective rights, for instance the conclusion of a contract of sale implies the legal power to transfer a property right.

In private law two generic types of rights are distinguished. First, the rights ad personam, or relative rights, that can only be invoked against specified other legal subjects. Such rights include those resulting from a contract, a tort action or unjustified enrichment. Second, the rights erga omnes, or absolute rights, that can be invoked against any and all legal subjects. Such rights include ownership, usufruct, right of way and intellectual property rights.

A claim right assumes an obligation or a duty on the side of one or more other legal subjects, e.g. a legal obligation to pay compensation (in the case of a tort or breach of contract), or a duty of non-interference (in the case of ownership).

A liberty right assumes that other legal subjects do not have a claim that one does or does not act in a specific way, e.g. in the case of ownership other legal subjects have no claim that the owner uses their property in a certain way, which demonstrates that property rights are bundles of claim and liberty rights.

A legal power assumes that one or more other legal subjects may be required to act or not act in a specific way, e.g. the legal power to transfer property implies that all legal subjects must now respect the right to property of the new owner and refrain from interference (in case of a property right), or the legal power of the government to impose taxes that implies that citizens must pay taxes (in case of the right of the state to unilaterally impose a duty to pay taxes).

The conceptual structure of modern positive law and the Rule of Law hinges on the institutional fact that only specified entities can act in law

In legal theory further distinctions are made, such as immunities, permissions and competences. The precise meaning of claims, liberties, powers, immunities, permissions and competences often differs between private and public law (and between national and international law), due to the different ways that the requirements of the Rule of Law play out (in private law the freedom and autonomy of private parties is foregrounded, whereas in public law the legality principle restricts the legal powers of the government). The applicable legal norms that constitute and limit subjective rights differ depending on jurisdiction, as it is positive law that decides on the attribution of subjective rights and positive law depends on jurisdiction.

Based on the discussion of foundational legal concepts in chapter 3, the concept of ‘legal powers’ can be situated in relation to the other foundational legal concepts as follows:

  • A legal power refers to the ability of a legal subject to achieve an intended legal effect, thus imposing legal obligations on other legal subjects to act or refrain from acting in a specific way. Legal powers are attributed to legal subjects by positive law, they are both constituted and limited by positive law. One can have a legal power to attribute legal powers. The legal norms that attribute legal powers may differ per jurisdiction and per domain of law, depending on the relevant sources of law.

The written or unwritten Constitution of a state attributes legal powers to legislate, administrate and adjudicate, thus calling them into existence (the constitution ‘constitutes’ these powers), and qualifying them (the constitution also ‘regulates’ these powers, e.g. by distributing them between countervailing powers). In the case of public law legal powers are constrained by the legality principle. This relates to the fact that the distribution of legal powers in a constitutional democracy takes into account the requirement of checks and balances of the Rule of Law. In the case of private law legal powers are constrained by the reasonableness principle (or equity in common law jurisdictions).

The attribution of legal power plays out in all domains of law. Private law, for instance, attributes to the owner of a legal good the legal power to transfer related property rights, provided specific conditions have been fulfilled. Criminal law, for instance, attributes to the court the legal power to impose specified (maximum) punishments, provided the conditions of a specific criminal offence have been fulfilled. Administrative law, for instance, attributes to legal subjects the legal power to object to decisions made by public administration, provided specific conditions apply. International law, for instance, attributes to states the legal power to conclude treaties, subject to the constraints imposed by the sources of international law.

Based on the discussion of the framing concepts in chapter 2, the concepts of ‘legal subjects, ‘subjective rights’ and ‘legal powers’ can be understood as follows:

  • The mode of existence of legal subjects, subjective rights and legal powers depends on the affordances of the text-driven normativity that informs modern positive law. As Diver notes in chapter 3, such text-driven normativity ‘cannot directly prevent [a] traffic accident by wresting control of the car from the driver, or by preventing her from driving at a certain speed — these would be examples of techno-regulation, compliance by design, or “legal by design”, each requiring technological intervention that goes beyond what text can impose. What text-driven normativity can do, however, is provide a framework that prefigures the relationships between those involved in and affected by that accident, and the consequences that flow from it after-the-fact.’ The conceptual structure of modern positive law and the Rule of Laws hinges on the institutional fact that specified entities can act in law, that is they can exercise their subjective rights and use their legal powers when navigating their material and institutional environment, thus co-shaping it by transferring rights and imposing legal obligations. The use of written and spoken speech acts in law affords a kind of legal protection that takes human agency seriously as legal subjects with rights and powers, taking note of the complex dynamics and myriad ambiguities that ‘make’ human society. Legal subjects must therefore not be naturalised, as if they are identical with the situated, embodied person they enable. As Diver suggests, legal subjectivity is like an avatar that allows human persons to engage with other legal subjects under cover of a web of legitimate expectations, thus providing a kind of protection that institutes the mode of existence of human persons as embodied agents that cannot be reduced to predictable and manipulable entities.

Based on the discussion of foundational legal concepts in chapter 3, the concepts of ‘legal reasoning’ and ‘legal interpretation’ can be situated in relation to the other foundational legal concepts as follows:

  • Legal reasoning concerns the justification of the determination of legal effect in a specific case. That legal effect could, for instance, be the attribution of a subjective right or a legal power. The justification requires both the selection and interpretation of the legally relevant facts in light of the applicable legal norm, and both the selection and interpretation of the applicable legal norm in light of the relevant facts.

The justification is provided in the form of a syllogism. The major is a legal norm that attributes specified legal effect if specified legal conditions are fulfilled, the minor is a specified set of legally relevant facts that, supported by evidence, fulfil the relevant legal conditions and the conclusion is the attribution of the specified legal effect. The syllogism that defines legal reasoning is not a method to find the legal effect but a way to test whether a legal norm does or does not apply. This test requires a decision on what is the applicable legal norm, what are the legally relevant facts and an act of interpretation as to the meaning of the norm and the meaning of the facts. The justification in the form of legal reasoning (the syllogism) assumes that the norm and the facts have been decided.

The following types of interpretation are deemed valid: ordinary meaning (grammatical or literal interpretation) based on the prevailing meaning of the norm’s text, framers’ intent (the intent of the legislature) as inferred from official documents, systematic interpretation based on the role the relevant norm plays in the context of the relevant legal system (its place in the relevant statute, its relationship with other norms whether higher norms such as a Constitution or Treaty or precedent) and teleological interpretation based on the purpose (telos) of the relevant legal norm, taking framers’ intent, ordinary meaning and systematic interpretation into account. Interpretation concerns positive law, as authoritatively decided by the legislature and the courts, taking into account the other sources of law, such as doctrine, legal custom and fundamental principles. This implies that interpretation concerns a specific jurisdiction and is not valid for other jurisdictions; what matters is not any interpretation but an interpretation that has legal effect within that jurisdiction. That is why legal reasoning is concerned with authoritative interpretation, anticipating how legal norms will be interpreted by the courts, highlighting that this is not merely about logic but about the legal power to decide a case. Under the Rule of Law that legal power ultimately rests with an independent judiciary and not with the legislature.

Speaking of the mode of existence of law elegantly avoids metaphysical discussions about the difference between the ontological and the epistemological nature of law

Legal reasoning is often defined as deontological reasoning (not about how things are but about how they should be). On top of that legal reasoning is qualified as non-monotonic reasoning and considered to involve defeasible logic, which means that whereas ‘if a then b’ is correct for now, additional information may render it incorrect. For instance, the legal norm ‘if a person has killed another person, that person is punishable’ may be correct, but its application may be incorrect due to a defence based on justification (e.g. self defence) or excuse (e.g. force majeure).

The syllogism requires interpretation of the legal norm in the light of the facts and interpretation of the facts in the light of the legal norm. It thus requires a decision about the extent to which a case is like or unlike other relevant cases. Such a decision requires one of two types of reasoning: either by analogy, arguing that since one case is like another the same legal norms applies to both or a contrario, arguing that since one case is different from another the same legal norm does not apply to both. The decision on whether a case is similar to a previously decided case must be argued and in a constitutional democracy the legal power to make that decision is with the courts, as required by the Rule of Law.

Based on the discussion of the framing concepts in chapter 2, the concepts of ‘legal subjects, ‘subjective rights’ and ‘legal powers’ can be understood as follows:

  • Speaking of the mode of existence of law elegantly avoids metaphysical discussions about the difference between ontological and epistemological nature of law. Instead, understanding ‘the way that law-as-we-know-it exists’ turns on a proper understanding of the performative effect of written and unwritten speech acts. In other words, engaging with the law requires key attention to what the ICI of the printing press affords in terms of ‘what law does’. As Duarte saliently writes in chapter 3: ‘[t]he activity of the judiciary is constituted and delimited by rules, principles, procedures and practices that, on the one hand, define its positive space of autonomy before other sovereign powers, and, on the other hand, negatively circumscribe its jurisdiction — i.e., its power to dictate the law — in the face of the spheres of legislature and administration. Legal reasoning incorporates the powers and constraints of the judiciary, thereby granting the legitimacy (indeed, the legality) of its authoritative discourse’. As law to ‘does’ just that ‘with words’, the kind of protection it offers cannot be understood as a matter of logical or causal inference and links the text-driven ‘design’ of legal protection directly to the Rule of Law. Interpretation. In the words of Duarte: ‘Legal reasoning affords a guarantee that the modes of veridiction of law apply to the judicial discourse’ and ‘By affording referential practice points that allow making sense of the legal system, interpretation affords legal protection in a double sense. One, it establishes a language game that constrains authoritative reasoning and interpretation, thereby preventing their fall into arbitrariness. Two, it settles the language game that equips the citizen to contest an administrative decision (or a judicial one, through appeal) under the terms of law’s mode of veridiction’.

If the outcome of code- or data-driven legal technologies does not, in itself, have legal effect, the deployment of those technologies may nevertheless impact the attribution of legal effect

If positive law attributes legal effect to the outcome of a code- or data-driven legal technology, that technology has intended legal effect. For instance if the legislature decides that an Act of Parliament is written simultaneously in computer code and in natural language, with both being enacted as representing the same set of legal norms, this example of Rules as Code itself has intended legal effect (e.g. a corporation may build its compliance software on the relevant computer code and claim it has thus achieved automated compliance). Or, if a court outsources the decision on what cases are inadmissible to an AI system that claims to predict their outcome, that outcome itself has legal effect (e.g. if the probability that the case will be won by whoever filed it is less than 5%, the case will be rejected as inadmissible). Or, if public administration outsources the decision on what applications for social security are granted to an ADM system, the output of that ADM system itself has legal effect (e.g. if the system has been certified as a proper implementation of all the relevant conditions for the application, its decision is final). Note that in all these examples legal remedies can be made available to enable people to contest the outcome: they could contest the interpretation that is inherent in the articulation of legislation in computer code; they could contest the reliability of the AI system used to predict the outcome of cases; and they could contest the software deployed to make decisions about social security. One of the many questions this raises is whether people are aware that a decision is made based on software and how, for instance, specific upstream design decisions end up having a major impact on the way they are treated.

If the outcome of code- or data-driven legal technologies does not, in itself, have legal effect, the deployment of those technologies may nevertheless impact the attribution of legal effect. This can be a matter of direct indirect impact.

If public administration deploys a programming language to transpose legislation into software and then uses that software to test whether people or corporations comply with the law, after which a civil servant decides whether or not a fine will be imposed, the software will have a direct effect on legal effect. Automation bias will contribute to civil servant semi-automatically following the output of the software, meaning that the situation comes close to the software itself having legal effect. Similarly, if a court deploys an AI system to predict the outcome of case applications, with the court registrar deciding whether or not a case is prioritised (e.g. in case of a heavy backlog of cases), the aim of this being more efficient implies that the predictions will have an indirect effect on the decisions made by the court, as their categorisation will push them up or down the line, causing a further delay for some cases though not for others. This will indirectly affect the legal effect, because such delay may cross the line towards indefinite postponements or because by the time the court comes to deal with the case other legislation may be in place or new precedent may be applied. A similar indirect effect on legal effect could play out when the police uses AI systems for crime mapping, shifting attention to particular neighbourhoods, redistribution their attention of, thus also changing their ability to detect crime in other neighbourhoods which will leave some offenders off the hook.

Though it may help to distinguish between direct and indirect effect on legal effect, the more important point is to distinguish between having legal effect and having effect on legal effect. In both cases the point will be to ensure legal protection in a way that avoids a crumbling effect on the checks and balances of the Rule of Law.

Legal effect is the vanishing point of modern positive law. Without it we have morality, politics or statistics but not law. Jurisdiction means nothing if there is no legal effect, the sources of law define what norms have legal effect, legal norms are defined by their legal effect and the attribution of legal subjectivity, subjective rights and legal powers all depend on their legal effect. The Rule of Law has no teeth and cannot offer any protection without legal effect.

Legal effect is neither causal nor logical but performative

Legal effect is neither causal nor logical but performative; the speech act that attributes legal effect does not cause or implicate that effect but affords the constitutive effect of ‘doing things with words’ in a very specific sense that is related to the internal and external sovereignty that grounds the ability to enforce legal norms. The ability to enforce, however, is not the physical or psychological cause of the effect – sovereignty itself is an institutional fact, i.e. the performative effect of a dynamic web of dedicated written and unwritten speech acts. Legal effect is text-driven, it is an affordance of the ICI of the printing press and cannot be reduced to physical force or mere convention.

Having legal effect is what defines modern positive law, or law-as-we-know-it. The fact that legal effect is a performative effect that ‘does what it says’ may seem a rather fragile basis for coordination, as the effect depends on a complex dynamic of linguistic interactions rather than brute force or strictly logical deductions. Brute force, however, would land us in a dictatorship, making us dependent on the wisdom or brutal self-interest of one person and their entourage; we invented the Rule of Law to overcome enlightened despotism, let alone ruthless tyranny. Strictly logical deduction or dynamic statistical prediction would, however, move us from human agency to its imitation, basically parasitising on our propositional representations or behavioural data by way of computational inferences. In the end, the performative effect of our complex interactional physical and institutional environment, is the mode of existence that best fits our embodied human agency, respecting our relational nature – provided we manage to institute, sustain and reinvent the check and balances that protect our relative autonomy (which is not a given but an affordance of how we rule ourselves).

Legislation is an institutional fact, built on written legal speech acts. It generates a specific type of legal certainty, as the text is equally valid for those subject to the relevant jurisdiction and available for contestation. Though some might think that contestation reduces legal certainty, the uncertainty that is inherent in human society means that contestation is a feature rather than a bug. Contestation enforces an agonistic debate about the interpretation of the text in light of the facts and vice versa, increasing the likelihood of a decision that takes all relevant arguments into account.

If positive law attributes legal effect to the outcome of a code- or data-driven legal technology, that technology has legal effect.

The key difference between having legal effect or having direct or indirect effect on legal effect resides in the different meanings of effect. Legal effect is a performative effect that does what it says, whereas the effect on legal effect is not a performative effect but a matter of influencing. In speech act theory this would be called a perlocutionary effect, that is an effect meant to bring about a certain situation.

Legal effect is a performative effect that does what it says, whereas the effect on legal effect is not a performative effect but a matter of influencing

Take the example of a marriage. When the civil registrar declares a couple husband and wife, they are not trying to influence them into getting married. The civil registrar is involved in a performative speech act; once spoken (or once added to the civil registry), the legal effects of marriage are instituted. The difference between the perlocutionary effect and the performative effect is immediately clear, because we all know that influencing people in the hope of them getting married will – in itself - not have any legal effect. However, we could imagine a situation where such influencing has direct or indirect effect on legal effect. For instance, arranged marriages or the use of marriage brokers may in a culture where arranged marriages are part of the local tradition, this may have an impact on the type of marriages that are being concluded as the brokers may favour marriages within one’s own cast – even if discrimination based on cast has been prohibited. If the civil registry deploys software to detect marriage fraud meant to obtain a residence permit for a person who would otherwise not be entitled to residence, the use of such software will influence the decisions of those working at the registry. The problem here is not only that the software may e.g. be biased against people with a specific ethnic background, but also the more fundamental fact that the software will work with some variable deemed relevant while leaving out others, thus surreptitiously influencing the kind of elements that affect decisions that may have legal effect.

Some scholars and policy makers, often influenced by the work of Kahneman regarding cognitive bias, believe that human beings are more prejudiced than machines, noting that the more intuitive motivation involved in human decision making is even less transparent than machine decisions, thanks to research in explainable data-driven AI or thanks to the fact that code-driven ADM can be traced back to ‘lossless’ implementation of a set of relevant rules. This could mean that all kinds of irrelevant, prejudicial and even racist motivations of individual human decision-makers will inevitably have a direct or indirect effect on legal effect, which would supposedly be very difficult to uncover compared to having software in place that is deemed objective as a matter of fact and whose reasons can be traced to a much further extent than the irrational biases of human beings.

We cannot assume that the meaning of legal effect as-we-know-it will not be affected by legal technologies that influence legal decision making

In this project we critique these assumptions from two perspectives. First, as to human bias, the work of e.g. Gadamer or Gigerenzer shows that bias is inevitable for decision making and not necessarily a bug but mostly a feature. Second, again as to human bias, it should be clear from this research study that the required justification of legal decision making restricts the decision space of public administration and courts, thus limiting unwarranted effect on legal effect on the side of human decision makers while nevertheless fostering the agency of the decision makers. In the upcoming research study on computational law we will investigate the claims made on behalf of legal technologies with regard to their potentially objective, unbiased nature as well as claims as to their transparency.

As a preparation for these investigations into the claims made on behalf of legal technologies and into the potential substantiation of those claims, we have developed an online webtool, called the Typology of Legal Technologies. This Typology can be found here and includes the vocabulary of foundational legal concepts that form the core of chapter 3 of this study and a vocabulary of similarly foundational computer science concepts, as well as a section on methodology, a section with answers to frequently asked questions and a section on dissemination. The Typology is meant as a mindset and a method, rather than merely a dataset or repository and we have been using it in teaching, tutorials at conferences and will be sharing it during upcoming expert meetings. The background research that informs the Typology will be further explored in the Research Study on Computational Law.

Finally, we cannot assume that the meaning of legal effect as-we-know-it today will not be affected by the deployment of legal technologies that influence the outcome of legal decision making. One of the reasons why we insist on adding inverted commas to ‘law’ when speaking of computational ‘law’ is that we do not consider indirect or direct effect on legal effect as itself being a matter of legal effect. However, once a legislature attributes legal effect to the outcome of these systems, even if under formal tutelage of a legal subject,

the use of the tech might drive a kind of conceptual slippage about the meaning of legal effect/the circumstances in which it obtains. this might go hand in hand with changes in practices/institutions considered to be ‘legal’ and need not arise out of the formal attribution of legal effect to these systems by law-as-we-know-it.1

Such slippage, though seemingly an indirect effect on legal effect, could paradoxically end up divulging the concept and practice of legal effect from its performative nature, thus opening the road to a rule by machines.2

References

  1. Pauline McBride in a comment to the conclusions. ↩

  2. On whether this makes sense, see Gianmarco Gori in his (yet unpublished) dissertation: https://flore.unifi.it/retrieve/e398c381-87d6-179a-e053-3705fe0a4cff/Law%2c%20Rules%2c%20Machines_Gianmarco%20Gori.pdf. ↩

This page was last updated on 4 January 2024.